Single crystal growth method, single crystal growth device and control apparatus therefor

Abstract

A single crystal growth method, a single crystal growth device and a control apparatus therefor. The method comprises: during a constant-diameter growth stage, denoting as D=F0(t) an initial function between the actual diameter of a crystal ingot and time t; when F0(t) satisfies a preset condition, locking the crystal pulling rate, and performing piecewise correction on the initial function by using measurement results of the actual diameter of the crystal ingot at subsequent n different measurement moments in sequence, so as to obtain a corrected piecewise function D=Fh(t); calculating F'h(t) at a corresponding moment; and comparing F'h(t) / (Ah*ωh) with a set value x and regulating the heating power in a single crystal furnace.

Description

Methods for single crystal growth, equipment for single crystal growth and control devices

[0001] Cross-reference to related applications

[0002] This application is based on and claims priority to Chinese Patent Application No. 202411505166.9, filed on October 25, 2024, the entire contents of which are incorporated herein by reference. Technical Field

[0003] This application relates to the field of crystal growth technology, and in particular to a method for single crystal growth, single crystal growth equipment and control device thereof. Background Technology

[0004] In the process of single-crystal constant-diameter growth, in order to grow a perfect crystal, it is necessary to lock the crystal growth rate, that is, to ensure that the crystal growth speed is the same as the set speed. Under the state of crystal growth rate lock, the crystal rod diameter is usually controlled by instantaneously changing the heating power or heating temperature, so as to control the crystal rod diameter through temperature pulses (or power pulses).

[0005] However, due to the thermal hysteresis of heat transfer in single crystal furnaces, accurate control is impossible during the regulation process, making it difficult to precisely control the diameter of the crystal rod. The method for controlling the diameter of the crystal rod in single crystal furnaces still needs improvement. Summary of the Invention

[0006] This application aims to at least solve one of the technical problems existing in the prior art. To this end, this application proposes a method for single crystal growth, a single crystal growth apparatus, and a control device thereof. The single crystal growth method corrects the initial function and utilizes F′ h (t) / (A h *ω h The heating power in the single crystal furnace is adjusted by comparing the value with the set value x, thereby achieving precise control of the crystal rod diameter.

[0007] The single crystal growth method according to the first aspect of this application includes: during the constant diameter growth stage, the initial function between the actual diameter of the crystal rod and time t is denoted as D = F0(t), where F0(t) = A0*sin(ω0*t+Φ)+d, or F0(t) = A0*cos(ω0*t+Φ)+d, d is the target diameter of the crystal rod, A0 is the initial amplitude, ω0 is the initial angular frequency, and Φ is the initial phase; when F0(t) meets a preset condition, the crystal pulling rate is locked, and the initial function is piecewise corrected using the measurement results of the actual diameter of the crystal rod at n different measurement times to obtain the corrected piecewise function D = F0(t) between the actual diameter of the crystal rod and time t. h (t), F h(t)=A h *sin(ω h *t+Φ)+d, or, F h (t)=A h *cos(ω h *t+Φ)+d, where A h The corrected amplitude, ω h The corrected angular frequency is D = F. h (t) includes several sub-functions: the function between the actual diameter of the crystal rod at time t0 and time t is the initial function D = F0(t); the calibrated function D = F1(t) at time t1; and t z The function D = F after time calibration z (t), t z The time is the z-th measurement time after the preset conditions are met, where z is a positive integer and 2≤z≤n; calculate F′ at the corresponding time. h (t), if F′ h (t) / (A h *ω h If F′ ≥ the set value x, then increase the heating power; if F′ h (t) / (A h *ω h )<0 and |F′ h (t) / (A h *ω h If |≥ set value x, then reduce the heating power, where A h ω h t h The amplitude and angular frequency at time x are set to 0. In two adjacent measurements, one of the parameters corresponding to the amplitude and angular frequency of the later measurement is equal to the corresponding parameter of the previous measurement, and the other is calculated to achieve a correction.

[0008] According to the single crystal growth method of this application embodiment, by continuously correcting the parameter amplitude and angular frequency in the initial function, the function model can be adjusted more accurately, which is beneficial to improving the accuracy of the corrected function and making it better fit the actual data. Compared with the initial function, the corrected function can more timely and accurately reflect the changing trend and rate of change of the actual diameter of the crystal rod at the current moment. Thus, the corrected function can regulate the crystal rod, making the actual change of the crystal rod in the constant diameter stage relatively less volatile, which is conducive to a more stable fluctuation of the actual diameter of the crystal rod and is conducive to growing a perfect crystal with uniform diameter.

[0009] In some embodiments, the initial function is piecewise corrected by sequentially using the measurement results of the actual diameter of the crystal rod at n different measurement times, including: when F0(t) meets the preset condition, the actual diameter D1 of the crystal rod is measured for the first time at time t1, and the corresponding parameter amplitude A1 is equal to the initial amplitude A0, and ω1 is calculated to achieve a correction; or, the corresponding parameter angular frequency ω1 is equal to the initial angular frequency ω0, and A1 is calculated to achieve a correction.

[0010] In some embodiments, in p consecutive corrections, at least one correction is performed to calculate the parameter amplitude, and at least one correction is performed to calculate the parameter angular frequency, where p is a positive integer and p≥2.

[0011] In some embodiments, in any two adjacent corrections, one correction calculates the amplitude of the parameter and the other correction calculates the angular frequency of the parameter.

[0012] In some embodiments, the constant diameter growth stage includes multiple sub-stages that proceed sequentially over time, with different set values ​​x corresponding to the multiple sub-stages, and the set values ​​x of the multiple sub-stages decreasing in chronological order.

[0013] In some embodiments, the current constant diameter length of the crystal rod is L, the target constant diameter length of the crystal rod is L', and satisfies the following condition: when 100mm ≤ L < 300mm, x = x1. When 300mm≤L<L'-500mm, x=x2. When L'-500mm≤L≤L', x=x3, 0.01≤x3≤1 / 2, where L'≥1300mm.

[0014] In some embodiments, when 300mm ≤ L < L' / 2, When L' / 2 ≤ L < L' - 500 mm

[0015] In some embodiments, when F0(t) = A0*sin(ω0*t+Φ)+d, the preset condition is F0(t) = d and F′0(t) > 0; when F0(t) = A0*cos(ω0*t+Φ)+d, the preset condition is F0(t) = d+A0.

[0016] In some embodiments, after F0(t) meets the preset conditions, a measurement is performed every preset time interval, where the preset time interval is less than 1 minute.

[0017] In some embodiments, if F′ h (t) / (A h *ω h If F′ ≥ the set value x, then the heating power is increased by ΔP1.h (t) / (A h *ω h ) < 0 and |F′ h (t) / (A h *ω h )| ≥ set value x, then reduce the heating power and the reduction amount is ΔP2, where ΔP1 = m1 * |F′ h (t) / (A h *ω h )|, ΔP2 = m2 * |F′ h (t) / (A h *ω h )|, both m1 and m2 are temperature control coefficients, 0 < m1 ≤ 15, 0 < m2 ≤ 15, where both m1 and m2 are positively correlated with ΔD, and ΔD is the difference in the diameter of the ingot between two adjacent measurement times.

[0018] In some embodiments, when it is necessary to adjust the heater power, the basic heating power of the heater adjusted each time is P0, calculate the period T of D = F h (t) h = 2π / ω h , if T h ≥ 40 min, then calibrate the basic heating power P0 of the heater so that P0 is equal to the average heating power within the preset time before the t h moment, where the value range of the preset time is T h / 8 ∼ T h / 2; if T h < 40 min, then keep P0 after the previous calibration unchanged.

[0019] In some embodiments, change the heating power through the top heater on the upper side of the crucible, m1 ≤ 1, m2 ≤ 1; or change the heating power through the side heater on the outer periphery of the crucible, 1 ≤ m1 ≤ 5, 1 ≤ m2 ≤ 5; or change the heating power through the bottom heater on the lower side of the crucible, 5 ≤ m1 ≤ 15, 5 ≤ m2 ≤ 15.

[0020] In some embodiments, when |F′ h (t) / (A h *ω h )| < set value x, if F h (t) > d, and F′ h (t) > 0, increase the actual liquid outlet distance; if F h (t) > d, and F′ h (t) < 0, reduce the actual liquid outlet distance; if F h (t) < d, and F′ h (t) < 0, reduce the actual liquid outlet distance; if F h(t) < d, and F′ h (t)>0, increase the actual liquid outlet distance.

[0021] In some embodiments, the change in the crucible rising rate is Δv, Δv=|k*ΔD / Δt|, where ΔD is the difference in the diameter of the crystal rod between two adjacent measurement moments, and Δt is the time interval between two adjacent measurement moments, where 0.1≤k≤2.

[0022] In some embodiments, the actual nozzle distance is taken as the target nozzle distance and fluctuates within the range of -5mm to 5mm.

[0023] The control device for a single crystal growth apparatus according to a second aspect embodiment of this application includes: a measuring mechanism, a data processing mechanism, a judgment mechanism, and a power adjustment mechanism. The measuring mechanism is used to measure the actual diameter D of the crystal rod. The data processing mechanism communicates with the measuring mechanism and is used to simulate an initial function D = F0(t) between the actual diameter of the crystal rod and time t, where F0(t) = A0*sin(ω0*t+Φ)+d, or F0(t) = A0*cos(ω0*t+Φ)+d, where d is the target diameter of the crystal rod, A0 is the initial amplitude, ω0 is the initial angular frequency, and Φ is the initial phase. The judgment mechanism is used to determine whether F0(t) meets a preset condition. When F0(t) meets the preset condition, the crystal pulling rate is locked, and the initial function is piecewise corrected using the measurement results of the actual diameter of the crystal rod at n different measurement times to obtain the corrected piecewise function D = F0(t) between the actual diameter of the crystal rod and time t. h (t), F h (t)=A h *sin(ω h *t+Φ)+d, or, F h (t)=A h *cos(ω h *t+Φ)+d, where A h The corrected amplitude, ω h The corrected angular frequency is D = F. h (t) includes several sub-functions: the function between the actual diameter of the crystal rod at time t0 and time t is the initial function D = F0(t); the calibrated function D = F1(t) at time t1; and t z The function D = F after time calibration z (t), t z The time is the z-th measurement time after the preset conditions are met, where z is a positive integer and 2≤z≤n. The data processing mechanism is used to ensure that, in two adjacent measurements, one of the corresponding parameters, amplitude and angular frequency, is equal to the corresponding parameter in the previous measurement, and the other is calculated to achieve a correction. The power adjustment mechanism is used to adjust the heating power, and the judgment mechanism is also used to judge F′.h (t) / (A h *ω h The relationship between F′ and the set value x, if F′ h (t) / (A h *ω h If F′ ≥ the set value x, then the power adjustment mechanism increases the heating power. h (t) / (A h *ω h )<0 and |F′ h (t) / (A h *ω h If |≥ set value x, then the power adjustment mechanism reduces the heating power.

[0024] A single crystal growth apparatus according to a third aspect of this application includes: a furnace body, a crucible, a heater, and a control mechanism. The crucible is disposed in the furnace body and defines a holding space. The heater is disposed in the furnace body and is used to heat the crucible. The control device is the control device of the single crystal growth apparatus according to the second aspect of this application described above. The power adjustment mechanism is used to adjust the power of the heater.

[0025] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0026] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which:

[0027] Figure 1 is a schematic flowchart of a single crystal growth method according to an embodiment of this application;

[0028] Figure 2 is a schematic flowchart of a single crystal growth method according to an embodiment of this application;

[0029] Figure 3 is a schematic flowchart of the heating power adjustment in a single crystal growth method according to an embodiment of this application;

[0030] Figure 4 is a simulation curve of the initial function between the actual diameter of the crystal rod and time under the preset conditions according to an embodiment of the present application. The solid line represents the function relationship before F0(t) satisfies the preset conditions, and the dashed line represents the corrected function after F0(t) satisfies the preset conditions.

[0031] Figure 5 is a corrected function relating the actual diameter of the crystal rod to time, and the derivative of the corrected function, according to an embodiment of this application.

[0032] Figure 6 is a corrected crystal rod diameter variation curve according to an embodiment of the present application, which incorporates liquid outlet distance control;

[0033] Figure 7 shows the fluctuation curve of the actual diameter of the lower crystal rod over time in the prior art;

[0034] Figure 8 is a curve showing the fluctuation of the actual diameter of the crystal rod over time, obtained by controlling the crystal rod diameter according to a corrected function, according to an embodiment of this application.

[0035] Figure 9 shows the fluctuation curves of the actual diameter of the crystal rod and the power of the heater over time in the prior art. Detailed Implementation

[0036] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described below can be combined with each other. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain this application, and should not be construed as limiting this application.

[0037] The following disclosure provides numerous different embodiments or examples for implementing various structures of this application. To simplify the disclosure, specific examples of components and arrangements are described below. These are merely examples and are not intended to limit the scope of this application. Furthermore, reference numerals and / or letters may be repeated in different examples. Such repetition is for simplification and clarity and does not in itself indicate a relationship between the various embodiments and / or arrangements discussed. In addition, various specific examples of processes and materials are provided in this application; however, those skilled in the art will recognize the applicability of other processes and / or the use of other materials.

[0038] In this application, unless otherwise specified, the following terms and symbols have the following meanings as defined below: D is the actual diameter of the crystal rod, t is time, D = F0(t) is the initial function between the actual diameter of the crystal rod and time, and D = F h (t) is the corrected function between the actual diameter of the crystal rod and time, d is the target diameter of the crystal rod, A is the amplitude, specifically the maximum difference between the actual diameter D and the target diameter d of the crystal rod within one period, ω is the angular frequency, Ф is the initial phase, A0 is the initial amplitude, ω0 is the initial angular frequency, x is the set value, ΔP1 and ΔP2 are the changes in heating power, L is the current constant diameter length of the crystal rod, m1 and m2 are the temperature control coefficients, and Δv is the change in the crucible rising rate. The term "perfect crystal" used in this paper does not mean an absolutely perfect crystal or a crystal without any defects, but rather allows for the presence of a very small number of one or more defects, insufficient to cause a significant change in a certain electrical or mechanical property of the crystal or the resulting wafer that would degrade the performance of the electronic devices made from it.

[0039] Hereinafter, with reference to the accompanying drawings, a method for single crystal growth according to an embodiment of this application will be described.

[0040] As shown in Figures 1 and 2, the method for single crystal growth includes: In the constant diameter growth stage, the initial function between the actual diameter of the crystal rod and time t is denoted as D = F0(t), where F0(t) = A0*sin(ω0*t+Φ)+d (as shown in Figure 4), or F0(t) = A0*cos(ω0*t+Φ)+d, where d is the target diameter of the crystal rod, A0 is the initial amplitude, ω0 is the initial angular frequency, Φ is the initial phase, and D is the actual diameter of the crystal rod. It can be seen that in the initial function F0(t), A0 > 0, and ω0 > 0.

[0041] When F0(t) meets the preset condition, the crystal pulling rate is locked, and the initial function is piecewise corrected using the measurement results of the actual diameter of the crystal rod at n different measurement times, so as to obtain the piecewise function D = F0(t) between the actual diameter of the crystal rod and time t. h (t), F h (t)=A h *sin(ω h *t+Φ)+d, or, F h (t)=A h *cos(ω h *t+Φ)+d, where A h The corrected amplitude, ω h Let be the corrected angular frequency, where D = F. h (t) includes several sub-functions: the function between the actual diameter of the crystal rod at time t0 and time t is the initial function D = F0(t); the calibrated function D = F1(t) at time t1; and t z The function D = F after time calibration z (t)(e.g., F) z (t)=A z *sin(ω z *t+Φ)+d, or, F z (t)=A z *cos(ω z *t+Φ)+d,)t z The time is the z-th measurement time after the preset conditions are met, where z is a positive integer and 2≤z≤n, and n is also a positive integer; calculate F′ at the corresponding time. h (t), if F′ h (t) / (A h *ω h If F′ ≥ the set value x, then increase the heating power; if F′ h (t) / (A h *ω h )<0 and |F′ h (t) / (A h *ω h )|≥Set value x (i.e. F′) h(t) / (A h *ω h If the absolute value of A is greater than or equal to the set value x), then reduce the heating power. h ω h t h The amplitude and angular frequency at time x are set to x > 0. Among two adjacent measurements, one of the parameters amplitude and angular frequency corresponding to the latter measurement is equal to the corresponding parameter of the former measurement, and the other is calculated to achieve a correction.

[0042] For example, taking the initial function as F0(t) = A0*sin(ω0*t+Φ)+d, where A0, ω0, and Φ are all known, and F0(t) satisfies the preset conditions, the measurement result obtained from the first measurement is the actual diameter of the crystal rod at time t1, which is D1. Using this first measurement result, the initial function is corrected to obtain the function D = F1(t) between the actual diameter D of the crystal rod at time t1 and time t. The measurement result obtained from the second measurement is the actual diameter D2 of the crystal rod at time t2. Using this second measurement result, the initial function is corrected to obtain the function D = F2(t) between the actual diameter D of the crystal rod at time t2 and time t. The measurement result obtained from the third measurement is the actual diameter D3 of the crystal rod at time t3. Using this third measurement result, the initial function is corrected to obtain the function D = F3(t) between the actual diameter D of the crystal rod at time t3 and time t. And so on, until the measurement result obtained from the z-th measurement is t. z The actual diameter of the crystal rod at this time is D z The initial function is corrected using the result of the z-th measurement to obtain t. z The function D = F, which relates the actual diameter D of the crystal rod at time t. z (t); ...; The measurement result obtained from the nth measurement is t n The actual diameter of the crystal rod at this time is D n The initial function is corrected using the results of the nth measurement to obtain t. n The function D = F, which relates the actual diameter D of the crystal rod at time t. n (t).

[0043] It can be seen that the above times t1, t2, ..., t n The times are all times corresponding to the actual diameter of the crystal rod in a single measurement, t z-1 Time and t z A time interval can correspond to two adjacent measurements. Therefore, the piecewise function D = F is obtained by sequentially using the results of n measurements to perform piecewise correction on the initial function. h (t) can include:

[0044] At time t1, the function D = F1(t) relating the actual diameter of the crystal rod to time t is, for example, F1(t) = A1*sin(ω1*t+Φ)+d;

[0045] At time t2, the function D = F2(t) relating the actual diameter of the crystal rod to time t is, for example, F2(t) = A2*sin(ω2*t+Φ)+d;

[0046] At time t3, the function D = F3(t) relating the actual diameter of the crystal rod to time t is, for example, F3(t) = A3*sin(ω3*t+Φ)+d;

[0047] …;

[0048] t z At time t, the function D = F z (t), for example F z (t)=A z *sin(ω z *t+Φ)+d;

[0049] …;

[0050] t n At time t, the function D = F n (t), for example F n (t)=A n *sin(ω n *t+Φ)+d.

[0051] Among them, A z >0, ω z If the value is greater than 0, considering that the parameter differences of the function are not significant in the short term between two adjacent measurements, one of the parameters corresponding to the amplitude and angular frequency in the later measurement is equal to the corresponding parameter in the previous measurement, while the other is calculated to achieve a correction. This facilitates the correction of the initial function and allows the corrected function to more accurately reflect the change of the actual diameter of the crystal rod over time. Therefore, by establishing an initial function and performing piecewise correction on the initial function using multiple actual measurement results after the initial function meets the preset conditions, the corrected function can more accurately reflect the change of the actual diameter of the crystal rod over time, facilitating more timely adjustment of the crystal rod diameter using the corrected function.

[0052] For example, for functions D = F1(t) and D = F2(t), there are two ways to correct the amplitude A2 and angular frequency ω2: First, take A2 = A1, and ω2 is calculated by substituting the actual diameter D2 of the crystal rod measured at time t2 into the function F2(t), that is, ω2 is calculated using the equation D2 = A2*sin(ω2*t+Φ)+d, where A2, D2, t2, Φ, and d are all known; Second, take ω2 = ω1, and A2 is calculated by substituting the actual diameter D2 of the crystal rod measured at time t2 into the function F2(t), that is, A2 is calculated using the equation D2 = A2*sin(ω2*t+Φ)+d, where ω2, D2, t2, Φ, and d are all known. Similarly, in any two subsequent adjacent measurements, for function F z (t)=A z *sin(ω z *t+Φ)+d and function F z+1 (t)=A z+1 *sin(ω z+1 For *t+Φ)+d: Ф is a constant, and the amplitude A z+1 and angular frequency ω z+1 There are two correction methods: First, take A. z+1 =A z , and ω z+1 By t z+1 The actual diameter D of the crystal rod measured at time 10:00 z+1 Substitute into function F z+1 ω is calculated from (t). z+1 Through equation D z+1 =A z+1 *sin(ω z+1 A is calculated by *t+Φ)+d. z+1 D z+1 t z+1 Both Ф and d are known; second, take ω. z+1 =ω z And A z+1 By t z+1 The actual diameter D of the crystal rod measured at time 1 z+1 Substitute into function F z+1 The result is calculated from (t), i.e., A z+1 Through equation D z+1 =A z+1 *sin(ω z+1 ω is calculated by *t+Φ)+d. z+1 D z+1 t z+1 Both Ф and d are known.

[0053] Then, using the corrected function, the corrected function F at the corresponding time point is calculated. h The first derivative of (t) F′ h(t), and according to F′ h (t) / (A h *ω h The heating power is dynamically adjusted by comparing the value of F′ with the set value x. h (t) / (A h *ω h If the actual diameter of the crystal rod is greater than or equal to the set value x, it indicates that the actual diameter of the crystal rod is in an upward phase and is increasing rapidly. Increasing the heating power at this point to suppress the increase in the actual diameter of the crystal rod is beneficial in causing the actual diameter of the crystal rod to fluctuate towards the target diameter. F′ h (t) / (A h *ω h )<0 and |F′ h (t) / (A h *ω h When the actual diameter of the crystal rod is greater than or equal to the set value x, it indicates that the actual diameter of the crystal rod is in the shrinking phase and is decreasing rapidly. Reducing the heating power at this time suppresses the decrease in the actual diameter of the crystal rod, which helps the actual diameter of the crystal rod fluctuate towards the target diameter. This results in a more stable change in the actual diameter of the crystal rod during the constant diameter growth stage in actual production, with a shorter fluctuation period. This improves the accuracy of controlling the actual diameter of the crystal rod and facilitates the maintenance of a stable V / G ratio during the constant diameter stage, thus contributing to the growth of perfectly uniform crystals. Therefore, the corrected function can achieve precise temperature control, which is helpful for the actual control of the crystal rod diameter.

[0054] For example, after correcting the initial function D = F0(t), we obtain the function D = F1(t) at time t1. At time t1 or between time t1 and time t2, we use F′1(t1) / (A1*ω1) to determine the adjustment of the heating power. F′4(t1) can be understood as the first derivative of function F1(t) at time t1, and compared to the initial function, it more accurately reflects the current trend of diameter change and the rate of change. After correction, we obtain the function D = F2(t) at time t2. We use F′2(t2) / (A2*ω2) to determine the adjustment of the heating power. F′2(t2) can be understood as the first derivative of function F2(t) at time t2, and compared to the initial function, it also more accurately reflects the current trend of diameter change and the rate of change. ...; After correction, we obtain the function D = F2(t) at time t2. z The function of time D = F z (t), using F′ z (t z ) / (A z *ω z To determine the adjustment of heating power, F′ z (t z This can be understood as a function F. z (t) at t zThe first derivative at time t is more accurate than the initial function in reflecting the changing trend and rate of change of the diameter at the current time; and so on. Then, calculate F′ at the corresponding time. h (t) can be understood as calculating F h (t) is the first derivative at the corresponding measurement time.

[0055] Of course, the embodiments of this application are not limited to this. In some other examples, for the time interval from t1 to t2 (i.e., t1≤t<t2), the corrected function is D=F1(t), and F′1(t1) / (A1*ω1) can be used to determine the adjustment of the heating power, or F′1(t1) / (A1*ω1) can be used. y1 The adjustment of heating power is determined by ) / (A1*ω1), F′1(t) y1 This can be understood as the function F1(t) at time t y1 The first derivative at time t y1 The time interval is any time between t1 and t2. Compared to the initial function, it can more accurately reflect the trend of diameter change and the rate of change at the current time. For times t2 to t3 (i.e., t2 ≤ t < t3), the corrected function is D = F2(t). F′2(t2) / (A2*ω2) can be used to determine the adjustment of heating power, and F′2(t2) / (A2*ω2) can also be used to determine the adjustment of heating power. y2 The adjustment of heating power is determined by ) / (A2*ω2), F′2(t) y2 This can be understood as the function F2(t) at time t y2 The first derivative at time t y2 The time interval is any time between t2 and t3. Compared to the initial function, it more accurately reflects the changing trend of the diameter at the current time and the rate of change; and so on. At this point, calculate F′ for the corresponding time. h (t) can be understood as calculating F h (t) is the first derivative of (t) at any time between two adjacent measurement times.

[0056] Of course, there are also some examples where t z Time to t z+1 Time (i.e., t) z ≤t<t z+1 The corrected functions are all D = F z (t), or F′ can be used z (t z ) / (A z *ω z This is used to determine the adjustment of heating power.

[0057] The examples above all use F0(t) = A0*sin(ω0*t+Φ)+d for correction and optimization. In the application of trigonometric functions, sine and cosine functions can be converted to each other through phase shift. Therefore, the derivation and optimization process of F0(t) = A0*cos(ω0*t+Φ)+d is similar to that of F0(t) = A0*sin(ω0*t+Φ)+d, and the detailed derivation process of F0(t) = A0*cos(ω0*t+Φ)+d will not be elaborated here.

[0058] It is understandable that after locking the crystal pulling rate, due to the complexity of the actual crystal growth situation, there is a certain difference between the actual change in the crystal rod diameter and the initial function. Specifically, under the locked pulling speed, when the heater uses temperature (or power) pulse control to control the diameter of the crystal rod, due to the thermal hysteresis of heat transfer in the single crystal furnace, if the pulse signal is too large and the duration is too long, it will lead to thermal shock and heat accumulation, which can easily cause diameter instability or even diameter loss of control, or even polycrystalline crystallization. On the other hand, if the pulse signal is too small and the duration is too short, it will not achieve the control effect. For example, the change of crystal rod diameter and heating power in some technologies is shown in Figure 9. The horizontal axis is the crystal rod constant diameter length, and the vertical axis is the measured crystal rod diameter. The target diameter of the crystal rod is 308 mm. When the constant diameter length of the crystal rod is 715 mm-725 mm, the actual diameter of the crystal rod is measured to be 310 mm, with a deviation of 2 mm. At this time, the heater performs a heating pulse. Due to the thermal hysteresis and the continuous growth of the crystal rod, the heat is applied to the position between the constant diameter length of the crystal rod of 735 mm-745 mm (the crystal growth rate is approximately 2 minutes for every 1 mm of length). This results in the actual measured diameter of 304 mm at the 745 mm position, with a deviation of 4 mm.

[0059] Therefore, the above-mentioned settings in this application perform piecewise correction on the initial function and use the corrected function as the basis for adjusting the heating power. This allows for more timely adjustment of the crystal rod diameter, reducing thermal hysteresis time and making the actual diameter fluctuation of the crystal rod more stable. As shown in Figures 7 and 8, Figure 7 is a graph showing the diameter fluctuation of the crystal rod over time during the actual growth process, obtained by adjusting the heating power according to a preset curve in the prior art. Figure 8 is a graph showing the diameter fluctuation of the crystal rod over time during the actual growth process, obtained by adjusting the heating power based on the preset curve in the prior art. It can be seen from the figures that the single crystal growth method of this application can significantly make the change in the actual diameter of the crystal rod during the constant diameter stage more stable, and the fluctuation period is relatively short, which is beneficial to improving the accuracy of controlling the actual diameter of the crystal rod.

[0060] According to the single crystal growth method of this application embodiment, by continuously correcting the parameter amplitude and angular frequency in the initial function, the function model can be adjusted more accurately, which is beneficial to improving the accuracy of the corrected function and making it better fit the actual data. Compared with the initial function, the corrected function can more timely and accurately reflect the changing trend and rate of change of the actual diameter of the crystal rod at the current moment. Thus, the corrected function can regulate the crystal rod, making the actual change of the crystal rod in the constant diameter stage relatively less volatile, which is conducive to a more stable fluctuation of the actual diameter of the crystal rod and is conducive to growing a perfect crystal with uniform diameter.

[0061] In some embodiments, during the constant diameter growth stage, the actual diameter of the crystal rod is continuously collected within a preset time period to obtain the change of the actual diameter D of the crystal rod with time t. At this time, multiple discrete data are obtained. By analyzing these discrete data, the range of diameter fluctuation can be obtained, and the periodic law of the change of the actual diameter of the crystal rod with time can also be obtained. By fitting a sine function or cosine function to these discrete data, the initial amplitude A0 and the initial angular frequency ω0 can be directly obtained from the fitting result, thereby obtaining A0 and ω0 that conform to the diameter change law, and thus simulating the initial function F0(t) from the multiple discrete data. The preset time period can be selected according to actual needs, such as the start of the constant diameter stage, which can be 1h to 2h, but is not limited to this. Of course, the preset time period can also end when the initial function fluctuates relatively evenly around the target diameter (i.e., the actual diameter of the crystal rod measured multiple times is partly greater than the target diameter and partly smaller than the target diameter), and then it is determined whether F0(t) meets the preset condition. Of course, the moment when F0(t) meets the preset condition can also be used as the end time of the preset time period.

[0062] When F0(t) satisfies the preset conditions, if the current time t0 corresponds to a single measurement time, that is, the actual diameter of the crystal rod is measured at time t0, or if the current time t0 does not correspond to a single measurement time, that is, the actual diameter of the crystal rod is not measured at time t0, then the actual diameter of the crystal rod corresponding to the current time t0 can be directly calculated by the initial function.

[0063] In some embodiments, the initial function is piecewise corrected by sequentially using the measurement results of the actual diameter of the crystal rod at n different measurement times. This includes: when F0(t) meets a preset condition, the actual diameter D1 of the crystal rod is measured for the first time at time t1, and the corresponding parameter amplitude A1 is equal to the initial amplitude A0. ω1 is calculated to achieve a first correction. In other embodiments, the initial function is piecewise corrected by sequentially using the measurement results of the actual diameter of the crystal rod at n different measurement times. This includes: when F0(t) meets a preset condition, the actual diameter D1 of the crystal rod is measured for the first time at time t1, and the corresponding angular frequency ω1 is equal to the initial angular frequency ω0. A1 is calculated to achieve a first correction.

[0064] Therefore, obtaining the function D = F1(t) relating the actual diameter of the crystal rod at time t1 to time t facilitates the simplification of function correction while more timely and accurately reflecting the change in the actual diameter of the crystal rod. It also facilitates the subsequent functions F2(t), F3(t), ..., F... n The acquisition of (t) can better and more accurately reflect the actual diameter change of the crystal rod, which is beneficial to enhancing the stability of the system.

[0065] In some embodiments, in p consecutive corrections, at least one correction calculates the parameter amplitude, and at least one correction calculates the parameter ω, where p is a positive integer and p≥2. That is, in p consecutive corrections, not all corrections calculate and correct the parameter amplitude, nor all corrections calculate and correct the parameter angular frequency. This helps to avoid increasing the error of one parameter by correcting only one parameter in p consecutive corrections, thus reducing error accumulation. It allows for more precise adjustment of the function model, improving the accuracy of the parameter amplitude and parameter angular frequency of the corrected function, enabling it to better fit the actual data, and making the corrected function more timely and accurate in reflecting changes in the crystal rod diameter compared to the initial function.

[0066] For example, the self-function F z (t)=A z *sin(ω z *t+Φ)+d to F z+p (t)=A z+p *sin(ω z+p For the p functions in *t+Φ)+d, after p consecutive corrections, for example, F z (t) to F z+2 (t) After two corrections, F z (t) to F z+3(t) After 3 corrections, in these p corrections, at least one correction is calculated to obtain the parameter amplitude by an equation, and at least one correction is calculated to obtain the parameter angular frequency by an equation. The number of corrections corresponding to the parameter amplitude calculated by an equation and the number of corrections corresponding to the parameter angular frequency calculated by an equation may be equal or unequal.

[0067] Taking the initial function F0(t)=A0*sin(ω0*t+Φ)+d as an example:

[0068] (1) If p = 2, the self-function F z (t)=A z *sin(ω z *t+Φ)+d to F z+2 (t)=A z+2 *sin(ω z+2 For *t+Φ)+d, one of the following correction methods can be used: First, the self-function F z (t) corrected to function F z+1 (t), take amplitude A z+1 =A z angular frequency ω z+1 By t z+1 The actual diameter D of the crystal rod measured at time 1 z+1 Substituting into the equation, we can calculate the independent function F. z+1 (t) corrected to function F z+2 (t), taking the angular frequency ω z+2 =ω z+1 , amplitude A z+2 By t z+2 The actual diameter D of the crystal rod measured at time 1 z+2 Substituting into the equation, we can calculate the result; secondly, the self-function F z (t) corrected to function F z+1 (t), take ω z+1 =ω z A z+1 By t z+1 The actual diameter D of the crystal rod measured at time 1 z+1 Substituting into the equation, we can calculate the independent function F. z+1 (t) corrected to function F z+2 (t), take amplitude A z+2 =A z+1 angular frequency ω z+2 By t z+2 The actual diameter D of the crystal rod measured at time 1 z+2 Substitute into the equation and calculate;

[0069] (2) If p = 3, the independent function F z (t)=A z *sin(ωz *t+Φ)+d to F z+3 (t)=A z+3 *sin(ω z+3 For *t+Φ)+d, one of the following correction methods can be used: First, the self-function F z (t) corrected to function F z+1 (t), take amplitude A z+1 =A z angular frequency ω z+1 By t z+1 The actual diameter D of the crystal rod measured at time 1 z+1 Substituting into the equation, we can calculate the independent function F. z+1 (t) corrected to function F z+2 (t), taking the angular frequency ω z+2 =ω z+1 , amplitude A z+2 By t z+2 The actual diameter D of the crystal rod measured at time 1 z+2 Substituting into the equation, we can calculate the independent function F. z+2 (t) corrected to function F z+3 (t), taking the angular frequency ω z+3 =ω z+2 , amplitude A z+3 By t z+3 The actual diameter D of the crystal rod measured at time 1 z+3 Substituting into the equation, we can calculate the result; secondly, the self-function F z (t) corrected to function F z+1 (t), take amplitude A z+1 =A z angular frequency ω z+1 By t z+1 The actual diameter D of the crystal rod measured at time 1 z+1 Substituting into the equation, we can calculate the independent function F. z+1 (t) corrected to function F z+2 (t), taking the angular frequency ω z+2 =ω z+1 , amplitude A z+2 By t z+2 The actual diameter D of the crystal rod measured at time 1 z+2 Substituting into the equation, we can calculate the independent function F. z+2 (t) corrected to function F z+3 (t), take amplitude A z+3 =A z+2 angular frequency ω z+3 By t z+3 The actual diameter D of the crystal rod measured at time 1 z+3 Substituting into the equation, we can calculate the result; third, the self-function F z (t) corrected to function Fz+1 (t), take amplitude A z+1 =A z angular frequency ω z+1 By t z+1 The actual diameter D of the crystal rod measured at time 1 z+1 Substituting into the equation, we can calculate the independent function F. z+1 (t) corrected to function F z+2 (t), take amplitude A z+2 =A z+1 angular frequency ω z+2 By t z+2 The actual diameter D of the crystal rod measured at time 1 z+2 Substituting into the equation, we can calculate the independent function F. z+2 (t) corrected to function F z+3 (t), taking the angular frequency ω z+3 =ω z+2 , amplitude A z+3 By t z+3 The actual diameter D of the crystal rod measured at time 1 z+3 Substituting into the equation, we can calculate the result; fourth, the self-function F z (t) corrected to function F z+1 (t), taking the angular frequency ω z+1 =ω z , amplitude A z+1 By t z+1 The actual diameter D of the crystal rod measured at time 1 z+1 Substituting into the equation, we can calculate the independent function F. z+1 (t) corrected to function F z+2 (t), take amplitude A z+2 =A z+1 angular frequency ω z+2 By t z+2 The actual diameter D of the crystal rod measured at time 1 z+2 Substituting into the equation, we can calculate the independent function F. z+2 (t) corrected to function F z+3 (t), take amplitude A z+3 =A z+2 angular frequency ω z+3 By t z+3 The actual diameter D of the crystal rod measured at time 1 z+3 Substituting into the equation, we can calculate the result; fifth, the self-function F z (t) corrected to function F z+1 (t), taking the angular frequency ω z+1 =ω z , amplitude A z+1 By t z+1 The actual diameter D of the crystal rod measured at time 1 z+1 Substituting into the equation, we can calculate the independent function F. z+1(t) corrected to function F z+2 (t), take amplitude A z+2 =A z+1 angular frequency ω z+2 By t z+2 The actual diameter D of the crystal rod measured at time 1 z+2 Substituting into the equation, we can calculate the independent function F. z+2 (t) corrected to function F z+3 (t), taking the angular frequency ω z+3 =ω z+2 , amplitude A z+3 By t z+3 The actual diameter D of the crystal rod measured at time 1 z+3 Substituting into the equation, we can calculate the result; sixth, the self-function F z (t) corrected to function F z+1 (t), taking the angular frequency ω z+1 =ω z , amplitude A z+1 By t z+1 The actual diameter D of the crystal rod measured at time 1 z+1 Substituting into the equation, we can calculate the independent function F. z+1 (t) corrected to function F z+2 (t), taking the angular frequency ω z+2 =ω z+1 , amplitude A z+2 By t z+2 The actual diameter D of the crystal rod measured at time 1 z+2 Substituting into the equation, we can calculate the independent function F. z+2 (t) corrected to function F z+3 (t), take amplitude A z+3 =A z+2 angular frequency ω z+3 By t z+3 The actual diameter D of the crystal rod measured at time 1 z+3 Substitute into the equation to calculate.

[0070] It's important to note that this only demonstrates a specific sequence of two and three consecutive calibration calculations. In practical applications, the order and number of calibrations can be adjusted based on specific data characteristics and optimization requirements. The key is to ensure that the parameter amplitude and angular frequency are adequately updated throughout the optimization process to improve the accuracy and reliability of the function model. In other words, p can be any positive integer greater than or equal to 2; for example, p can also be 4, 5, or higher.

[0071] In some embodiments, in any two adjacent corrections, one correction calculates the parameter amplitude, and the other calculates the parameter angular frequency. By alternately fixing and updating the parameter amplitude and parameter angular frequency, the corrected function can more closely approximate the actual crystal growth process. Each iteration is fine-tuned based on the result of the previous iteration, thereby reducing the accumulation of errors and improving the reliability of the function.

[0072] For example, the parameter obtained from the previous correction, either the amplitude parameter or the angular frequency parameter, is directly substituted into the next correction calculation. That is, for the function F... z (t)=A z *sin(ω z *t+Φ)+d、function F z+1 (t)=A z+1 *sin(ω z+1 *t+Φ)+d and function F z+2 (t)=A z+2 *sin(ω z+2 Regarding *t+Φ)+d: If the amplitude A z+1 =A z angular frequency ω z+1 Calculations show that the angular frequency ω is taken as... z+2 =ω z+1 , amplitude A z+2 Calculations show that if the angular frequency ω z+1 =ω z , amplitude A z+1 Calculations show that the amplitude A is taken as the result. z+2 =A z+1 angular frequency ω z+2 The calculations show that whether the amplitude parameter or the angular frequency parameter is substituted first for correction calculation, as long as the principle of alternating updates is maintained and each update is based on the latest measurement data and the results of the previous iteration, it will help to further improve the reliability and accuracy of the corrected function.

[0073] In some embodiments, the constant diameter growth stage includes multiple sub-stages that proceed sequentially over time. The setpoints x for these sub-stages are unequal, and the setpoints x decrease sequentially over time. During the constant diameter growth stage of the crystal rod, the thermal hysteresis time changes accordingly as the constant diameter length gradually increases; the longer the constant diameter length, the shorter the thermal hysteresis time. To accurately match this dynamic change, an intelligent control strategy is adopted. This involves subdividing the growth into multiple sub-stages based on different ranges of constant diameter length, and setting a unique setpoint x for each sub-stage. Specifically, the temperature pulse activation conditions differ for each sub-stage, and the temperature pulse occurrence conditions differ for each region. These setpoints are adjusted progressively with increasing constant diameter length, allowing for flexible adjustment of heating conditions to adapt to changes in thermal hysteresis time. This ensures the stability of the crystal rod growth process and achieves more precise heating control.

[0074] In some embodiments, the current constant diameter length of the crystal rod is L, the target constant diameter length of the crystal rod is L', and when 100mm ≤ L < 300mm, x = x1. When 300mm≤L<L'-500mm, x=x2. When L'-500mm≤L≤L', x=x3, 0.01≤x3≤1 / 2, where L'≥1300mm. The initial function can be either a sine or cosine function.

[0075] For example, assuming the target constant diameter length of the crystal rod L' = 2000 mm, and the actual constant diameter length of the crystal rod is between 0 and L < 100 mm, in the initial stage of constant diameter crystallization, the protective gas (e.g., argon) and water cooling jacket in the crystal growth equipment have a significant impact on the crystal, resulting in large fluctuations in the actual diameter of the crystal rod. At this time, it is difficult to lock the crystal pulling rate. In the initial stage of constant diameter crystallization, when the current constant diameter length L of the crystal rod is between 100 mm and L < 300 mm, x = x1. In the initial stage of a constant-diameter process, there is a large amount of liquid molten material, resulting in a longer thermal hysteresis time. Therefore, the temperature response to changes in heating power is relatively delayed. Consequently, temperature pulses should not be too long, requiring shorter temperature control times. At this point, |F′ h (t) / (A h *ω h Temperature control is activated when |≥x1, i.e., F′ h (t) / (A h *ω h When )≥x1, increase the heating power, while F′ h (t) / (A h *ω h )<0 and |F′ h (t) / (A h *ω hIf x ≥ x1, reduce the heating power; as the constant diameter length increases, the growth tends to stabilize. In the middle stage of constant diameter growth, when the current constant diameter length of the crystal rod is between 300mm and L < 1500mm, the thermal hysteresis time is relatively shortened, x = x2. At this moment |F′ h (t) / (A h *ω h Temperature control is activated when | ≥ x2, i.e., F′ h (t) / (A h *ω h When )≥x2, increase the heating power, while F′ h (t) / (A h *ω h )<0 and |F′ h (t) / (A h *ω h When |≥x2, reduce the heating power; in the later stage of constant diameter, when the current constant diameter length of the crystal rod is between 1500mm and 2000mm, the melt volume is further reduced, the remaining melt volume is small, the thermal hysteresis time is short, and the temperature control intervention time can be extended. At this time, x=x3, 0.01≤x3≤1 / 2, and |F′ h (t) / (A h *ω h )|≥x3 activates temperature control, i.e., F′ h (t) / (A h *ω h When )≥x3, increase the heating power, while F′ h (t) / (A h *ω h )<0 and |F′ h (t) / (A h *ω h When x ≥ x3, the heating power is reduced; this allows for a longer temperature control time without causing excessive heat accumulation. By changing the constant diameter length of the crystal rod and setting the corresponding set value, intelligent adjustment of the temperature control conditions is achieved, effectively avoiding the adverse effects that heat accumulation may cause. It can be understood that during the constant diameter growth stage, changing the heating power can be understood as applying a temperature pulse, and the duration of the temperature pulse can be understood as the duration after changing the heating power and maintaining it.

[0076] Of course, the constant diameter growth stage can be further divided into two, three, or more sub-stages. As the crystal rod grows, the temperature control conditions change to adapt to the changes in crystal rod growth.

[0077] In some embodiments, when 300mm ≤ L < L' / 2, When L' / 2 ≤ L < L' - 500 mm It can be seen that during the mid-stage of equal diameter growth, the growth tends to stabilize. When the length of the equal diameter is 300mm ≤ L < L' - 500mm, x = x2. x2 can be further subdivided to improve the quality and growth efficiency of the crystal rod.

[0078] For example, assuming the target constant diameter length of the crystal rod is L' = 2000 mm, and the current constant diameter length of the crystal rod is between 300 ≤ L < 1000 mm, the thermal hysteresis time is relatively shortened, x = x2. At this moment |F′ h (t) / (A h *ω h Temperature control is activated when | ≥ x2, i.e., F′ h (t) / (A h *ω h When )≥x2, increase the heating power, while F′ h (t) / (A h *ω h )<0 and |F′ h (t) / (A h *ω h When x ≥ x2, reduce the heating power; as the constant diameter length increases, the growth further tends to stabilize. The current constant diameter length of the crystal rod is between 1000mm and 1500mm. In this sub-stage, the crystal rod diameter tends to stabilize further, the melt stability decreases, and the thermal hysteresis time becomes shorter. x = x2. |F′ h (t) / (A h *ω h )|≥x2 activates temperature control, i.e., F′ h (t) / (A h *ω h When )≥x2, increase the heating power, while F′ h (t) / (A h *ω h )<0 and |F′ h (t) / (A h *ω h When L ≥ x2, the heating power is reduced. By further subdividing the range of x2, the heating power can be adjusted more precisely to adapt to minute changes during the crystal growth process, which helps to reduce fluctuations during crystal growth and thus improve the uniformity and consistency of the crystal. Optionally, the time period corresponding to L < L' / 10 can be used as the preset time period described throughout the text to construct the initial function.

[0079] In some embodiments, when F0(t) = A0*sin(ω0*t+Φ)+d, the preset condition is F0(t) = d and F′0(t) > 0; when F0(t) = A0*cos(ω0*t+Φ)+d, the preset condition is F0(t) = d+A0. It can be seen that when F0(t) = A0*sin(ω0*t+Φ)+d, the preset condition is F0(t) = d and F′0(t) > 0, ensuring that the actual diameter D of the crystal rod at a certain time t is equal to the target diameter d, and the initial phase is 0, i.e., Φ = 0. At this time, as shown in Figure 4, the initial function between the actual diameter of the crystal rod and time is D = F0(t), F0(t) = A0*sin(ω0*t)+d. Moreover, the time satisfying the preset condition indicates that the actual diameter D of the crystal rod is increasing with time t, which facilitates correction from the initial position of one fluctuation cycle of the initial function, simplifying the correction and reducing the computational difficulty; F When F0(t) = A0*cos(ω0*t+Φ)+d, the preset condition is F0(t) = d+A0. This ensures that the actual diameter of the crystal rod at the moment the preset condition is met is equal to the maximum fluctuation relative to the target diameter d. At this point, the initial function is at its peak, with an initial phase of 0, i.e., Φ = 0, or F0(t) = A0*cos(ω0*t)+d. Subsequent corrections can be made to the parameter amplitude and angular frequency. The simplified initial function makes subsequent data processing and correction easier to operate, and it facilitates correction from the initial position of one fluctuation cycle of the initial function, simplifying the correction and reducing computational difficulty. Clearly, the above settings help reduce errors that are easily introduced into subsequent correction processes due to the initial phase Φ≠0, thus further improving correction accuracy.

[0080] Of course, the preset conditions can also be other conditions. Under other conditions, the initial phase of the initial function is not 0, i.e., Ф≠0, and subsequent correction calculations of the initial function can still be performed. Optionally, the initial phase Ф can be calculated when simulating the initial function D=F0(t), or the initial phase Ф can also be calculated from the actual diameter of the crystal rod at the current moment when the preset conditions are met. Regardless of which method is used to obtain the initial phase Ф, Ф can be 0 or not equal to 0.

[0081] In some embodiments, after F(t) meets the preset conditions, a measurement is performed every preset time interval. The preset time interval is less than 1 minute. During the crystal growth process, the crystal growth rate is generally around 1 mm / min. Therefore, the preset time interval between adjacent measurements is less than 1 minute so that the time interval between adjacent measurements has good compatibility with the crystal growth rate. It is not easy for the corresponding parameters between two measurements to accumulate large errors due to excessively long measurement intervals. These errors will gradually accumulate and affect the accuracy of the correction function, the control precision of the system, and thus the quality of the final product. High-frequency measurement helps to reduce such error accumulation and improve the stability and reliability of the system.

[0082] Optionally, the preset time is less than 10 s, which can capture subtle changes in diameter more promptly, enabling the system to respond and adjust more quickly. With the increase in measurement frequency, the system can obtain more real-time data points for calibration calculations and make more precise control adjustments accordingly, thereby reducing the fluctuation range of the actual diameter of the crystal bar.

[0083] Furthermore, when the preset time is 1 s, real-time monitoring of the growth process is achieved, and any tiny change in the diameter of the crystal bar can be immediately captured. Through an almost continuous data stream, the system can more accurately evaluate the change trend of the diameter and make finer adjustments to maintain the stability of the diameter. The high-frequency measurement and rapid feedback mechanism contribute to enhancing the overall stability of the system and reducing fluctuations and deviations caused by external interference.

[0084] In some embodiments, if F′ h (t) / (A h *ω h ) ≥ the set value x, then increase the heating power by an amount of ΔP1; if F′ h (t) / (A h *ω h ) < 0 and |F′ h (t) / (A h *ω h )| ≥ the set value x, then decrease the heating power by an amount of ΔP2. Here, ΔP1 = m1 * |F′ h (t) / (A h *ω h )|, ΔP2 = m2 * |F′ h (t) / (A h *ω h )|, both m1 and m2 are temperature control coefficients, 0 < m1 ≤ 15, 0 < m2 ≤ 15. Due to differences in the heat field insulation performance, heater capabilities, etc., the temperature control coefficients are different. For example, for a heat field with poor insulation performance, a larger temperature control coefficient is required so that more heat can act on the melt.

[0085] It can be seen that ΔP1 = m1 * |F′ h (t) / (A h *ω h )| represents the amount of heating power that the system needs to increase when the diameter is increasing, and ΔP2 = m2 * |F′ h (t) / (A h *ω hThe value () indicates the amount of heating power the system needs to reduce when the diameter is decreasing. The increase in heating power is positively correlated with the actual diameter growth rate of the crystal rod, and the decrease in heating power is positively correlated with the actual diameter decrease rate of the crystal rod. By linking the adjustment of heating power to the absolute value of the diameter change rate, it is easier to match the temperature pulse with the fluctuation of the actual diameter of the crystal rod. The system can more accurately control the stability of the diameter. When the diameter deviates from the target value, the system can immediately adjust the heating power according to the degree and direction of the deviation, thereby quickly restoring the crystal rod to a constant diameter growth state. This helps to reduce the diameter fluctuation range and improve the stability and consistency of the growth process.

[0086] It is understandable that the heating power can be adjusted by any one or more heaters in the single crystal growth equipment.

[0087] For example, both m1 and m2 are positively correlated with ΔD, where ΔD is the difference in the diameter of the crystal rod at two adjacent measurement times, ΔD = D z+1 -D z If ΔD is larger, then m1 and m2 will be larger, and the change in heating power will be larger. This will help to achieve a better match between the temperature pulse and the actual diameter fluctuation of the crystal rod, and further improve the stability of the control over the crystal rod diameter.

[0088] In some embodiments, the single crystal growth method further includes: the base heating power of the heater adjusted each time is P0. It can be understood that the base heating power of the heater adjusted each time is P0, meaning that each time the heater power is adjusted, it is based on the base heating power P0, for example, by increasing or decreasing the base heating power P0, i.e., using the base heating power P0 as the adjustment reference; for example, if the heating power is increased by ΔP1 each time the heating power is adjusted, the adjusted heater power is P0 + ΔP1; similarly, if the heating power is decreased by ΔP2 each time the heating power is adjusted, the adjusted heater power is P0 - ΔP2.

[0089] The method for single crystal growth also includes: when the heater power needs to be adjusted, the base heating power of the heater for each adjustment is P0. Calculate D = F. h The period T of (t) h =2π / ω h If T h If the heating power adjustment takes ≥40 minutes, it indicates that the effect is slight to negligible, and the basic heating power P0 is insufficient to meet the actual needs. Therefore, the basic heating power P0 of the heater should be calibrated so that the calibrated P0 equals t. h The average heating power over a preset time period prior to the specified time, where the preset time ranges from T. h / 8~T h / 2, so that the adjusted base heating power can be based on t h The thermal history is obtained within a preset time period before time t, which makes it easier to obtain the thermal history within time t. h After adjusting the heater power based on the corrected base heating power P0, the crystal rod diameter can quickly converge, ensuring that the heating power adjustment achieves the desired purpose; if T h If the surface heating power adjustment effect meets expectations within 40 minutes, then the P0 value after the previous calibration remains unchanged, and the adjustment reference for the heating power at this time is the P0 value after the previous calibration.

[0090] This is understandable; let's denote the preset time as T. h ', due to the preset time T h The range of values ​​for ' is T. h / 8~T h / 2, i.e., T h / 8≤T h '≤T h / 2,t h The preset time before the time is (t) h -T h ') Time to t h The corresponding time interval between moments; for example, t h The time is 3:40, and the preset time is T. h If ' is 40 minutes, then t h The preset time before the time corresponds to the period between 3:00 and 3:40. The corrected P0 is equal to the average heating power between 3:00 and 3:40.

[0091] For example, calculate F′1(t1) at time t1. If F′1(t1) / (A1*ω1)≥ set value x, increase the heating power by ΔP1 and calculate the period T1=2π / ω1 of D=F1(t). If T1≥40min, calibrate the basic heating power P0 of the heater so that the corrected P0 is equal to the average heating power in the preset time before time t1 (i.e., the corrected P0 is equal to the average heating power between time (t1-T1') and time t1, T1 / 8≤T1'≤T1 / 2). The adjusted heater power is the sum of ΔP1 and the corrected P0. If T1<40min, keep the previously calibrated P0 unchanged. If F′1(t) / (A1*ω1)<0 and |F′1(t) / (A1*ω1)|≥set value x, then reduce the heating power by ΔP2, and calculate the period T1=2π / ω1 of D=F1(t); if T1≥40min, calibrate the basic heating power P0 of the heater so that the corrected P0 is equal to the average heating power in the preset time before time t1 (i.e., the corrected P0 is equal to the average heating power between time (t1-T1') and time t1, T1 / 8≤T1'≤T1 / 2), and the adjusted heater power is the corrected P0 minus ΔP2; if T1<40min, then keep the previously calibrated P0 unchanged.

[0092] Next, calculate F′2(t2) at time t2. If F′2(t2) / (A2*ω2) ≥ the set value x, then increase the heating power by ΔP1, and calculate D = F. \2 The period of (t) is T2 = 2π / ω2; if T2 ≥ 40 min, calibrate the basic heating power P0 of the heater so that the corrected P0 is equal to the average heating power within the preset time before time t2 (i.e., the corrected P0 is equal to the average heating power between time (t2-T2') and time t2, T2 / 8 ≤ T2' ≤ T2 / 2), and the adjusted heater power is the sum of ΔP1 and the corrected P0; if T2 < 40 min, then keep the previously calibrated P0 unchanged. If F′2(t) / (A2*ω2) < 0 and |F′2(t) / (A2*ω2)| ≥ the set value x, then reduce the heating power by ΔP2, and calculate D = F \2 The period of (t) is T2 = 2π / ω2; if T2 ≥ 40 min, calibrate the basic heating power P0 of the heater so that the corrected P0 is equal to the average heating power within a preset time before time t2 (i.e., the corrected P0 is equal to the average heating power between time (t2-T2') and time t2, T2 / 8 ≤ T2' ≤ T2 / 2), and the adjusted heater power is the corrected P0 minus ΔP2; if T2 < 40 min, then keep the calibrated P0 at time t1 unchanged. This process is repeated. It can be understood that T1', T2', ..., T h'、... etc. can be equal, or at least two can be unequal.

[0093] In some embodiments, the heating power is changed by the top heater on the upper side of the crucible, where m1≤1 and m2≤1, for example, m1 and m2 are 0.2, 0.5, 0.7, or 1; or, the heating power is changed by the side heater on the outer periphery of the crucible, where 1≤m1≤5 and 1≤m2≤5, for example, m1 and m2 are 1.2, 2.5, 3.7, 4.3, or 5; or, the heating power is changed by the bottom heater on the lower side of the crucible, where 5≤m1≤15 and 5≤m2≤15, for example, m1 and m2 are 5.8, 6.5, 7.7, 8.3, 9, 11, 13, 14, or 15. When the same temperature pulse is required, different heaters are positioned differently and spaced differently from the crystal rod. Consequently, different heaters have different thermal hysteresis. By varying the heating power of different heaters, the actual needs of the crystal rod can be met. Different heaters are set with different temperature control coefficients, which makes it easier for the temperature pulse generated by changing the heating power of the corresponding heater to better match the temperature pulse required to control the actual diameter change of the crystal rod. The system can select the most suitable heater and adjustment strategy based on the current actual diameter and the target diameter of the crystal rod. This flexibility helps to optimize the growth process and improve the quality and production efficiency of the crystal rod.

[0094] It is understandable that if the single crystal growth equipment includes a top heater, the top heater is relatively close to the molten solid-liquid interface within the crucible. The top heater plays a dominant role in adjusting the heating power, ensuring efficient heat transfer and directly influencing the overall growth process of the crystal rod. Conversely, if the single crystal growth equipment includes a bottom heater, the bottom heater is relatively far from the molten solid-liquid interface within the crucible. Therefore, in this application, the temperature control coefficients of different heaters are matched and correspond to each other to meet the crystal rod control requirements.

[0095] It is understandable that, in the process of controlling the diameter of the crystal rod, as long as the corresponding heating power can be met, one of these three heaters, or a random combination of these three heaters, can be used to work together to regulate the temperature change without affecting the growth process of the crystal rod.

[0096] In some embodiments, as shown in FIG2, |F′ h (t) / (A h *ω h When the actual diameter of the crystal rod changes at a rate less than the set value x, the diameter is basically in the peak region. To prevent diameter deviation, the diameter can be controlled by the liquid outlet distance, which can be adjusted in several different ways:

[0097] First, if F h(t)>d, and F′ h (t)>0, increase the actual liquid nozzle distance. At this time, the actual diameter of the crystal rod is greater than the target diameter of the crystal rod, the rate of change of the actual diameter of the crystal rod is positive, the actual diameter of the crystal rod is still increasing, and it is basically in the peak diameter region. At this time, the actual liquid nozzle distance can be increased to change the rate of change of the actual diameter of the crystal rod, so that the rate of change of the actual diameter of the crystal rod can be reduced from a positive value or changed to a negative value. The actual diameter of the crystal rod gradually decreases and approaches the target diameter of the crystal rod, thereby ensuring the stability of the crystal rod diameter in the production process.

[0098] Second, if F h (t)>d, and F′ h When F(t) < 0, the actual liquid nozzle distance is reduced. At this time, the actual diameter of the crystal rod is greater than the target diameter of the crystal rod, and the rate of change of the actual diameter of the crystal rod is negative. The actual diameter of the crystal rod is constantly decreasing. By reducing the actual liquid nozzle distance, the rate of change of the actual diameter of the crystal rod is changed, so that the actual diameter of the crystal rod is not easily reduced to be too small than the target diameter of the crystal rod. For example, by simultaneously adjusting the rising rate of the crucible and the pulling rate of the seed crystal, the system can achieve fine control of the growth rate and diameter of the crystal rod, ensuring that the actual diameter of the crystal rod can gradually approach and stabilize within the range of the target diameter of the crystal rod. The pulling rate of the seed crystal is the pulling rate of the crystal. The initial adjustment of the pulling rate of the seed crystal can be based on the crystal pulling rate locked when F(t) meets the preset conditions. The subsequent adjustment of the pulling rate of the seed crystal can be adjusted based on the current moment.

[0099] It is understandable that in the process of reducing the actual liquid outlet distance, in order to avoid the molten liquid level in the crucible from exceeding the solid-liquid interface, the above-mentioned settings in this application change the seed crystal pulling rate so that the molten liquid level will not exceed the solid-liquid interface while reducing the actual liquid outlet distance, which helps to reduce the risk of crystal rod breakage.

[0100] Third, if F h (t) < d, and F′ h (t) < 0, reduce the actual liquid outlet distance. At this time, the actual diameter of the crystal rod is smaller than the target diameter of the crystal rod, and the rate of change of the actual diameter of the crystal rod is negative. The actual diameter of the crystal rod is constantly decreasing. By reducing the actual liquid outlet distance, the rate of change of the actual diameter of the crystal rod is changed. The rate of change of the actual diameter of the crystal rod increases from a negative value to a positive value, so the actual diameter of the crystal rod is constantly increasing and gradually approaches the target diameter of the crystal rod, ensuring the stability of the crystal rod diameter in the production process.

[0101] Fourth, if F h (t) < d, and F′ h(t)>0, increase the actual liquid outlet distance. At this time, the actual diameter of the crystal rod is smaller than the target diameter of the crystal rod. The rate of change of the actual diameter of the crystal rod is positive, and the actual diameter of the crystal rod is constantly increasing. By increasing the actual liquid outlet distance, the rate of change of the actual diameter of the crystal rod is reduced, which makes it easier for the actual diameter of the crystal rod to continuously increase, and it is not easy to increase excessively to the target diameter of the crystal rod, thereby ensuring the stability of the crystal rod diameter in the production process.

[0102] The liquid outlet distance is the interval between the lower end of the guide tube and the solid-liquid interface. By adjusting the liquid outlet distance, the temperature gradient in the melt can be adjusted, thereby affecting the growth rate of the crystal rod.

[0103] It is evident that the liquid outlet distance can be altered by adjusting the crucible's rising rate and the seed crystal's pulling rate. This allows for dual intervention of temperature pulse control and liquid outlet distance control, ensuring that the actual crystal rod diameter remains within the target diameter range throughout the growth process, while minimizing the amplitude and period of fluctuation and exhibiting a rapid control response. Furthermore, by simultaneously adjusting the crucible's rising rate and the seed crystal's pulling rate, the relative position of the crystal rod to the molten surface remains constant.

[0104] It is understood that the "changing the crucible rising rate" and "changing the seed crystal pulling rate" in the above-mentioned scheme of this application are both adjusted directly based on the current moment. For example, if the crucible rising rate at the current moment is v, and it is necessary to increase the crucible rising rate by Δv, then the crucible rising rate is directly increased from v to the target rising rate v + Δv. Similarly, if it is necessary to decrease the crucible rising rate by Δv, then the crucible rising rate is directly decreased from v to the target rising rate v - Δv. Obviously, when v - Δv > 0, the crucible is still in the rising state; if v - Δv < 0, the crucible is in the falling state. In addition, when the liquid nozzle distance is used to control the actual diameter of the crystal rod, the heating power can be adjusted according to a preset curve. The setting of the preset curve is well known to those skilled in the art and will not be described in detail here.

[0105] Optionally, F′ h The relationship between F′ and 0 can be determined by the results of two consecutive measurements. If the actual diameter of the crystal rod obtained in the later measurement is greater than the actual diameter obtained in the earlier measurement, then F′ h If (t) > 0, and the actual diameter of the crystal rod obtained in the later measurement is greater than the actual diameter of the crystal rod obtained in the previous measurement, then F′ h (t) < 0. Of course, you can also determine this by directly differentiating the corrected function.

[0106] In some embodiments, in the single crystal growth method, the heating power may be adjusted without adjusting the actual liquid nozzle distance; or, both the heating power and the actual liquid nozzle distance may be adjusted. It is understood that, compared with the method of "single adjustment of heating power" (adjusting the heating power without adjusting the actual liquid nozzle distance), the method of "adjusting both the heating power and the actual liquid nozzle distance" is beneficial to reducing the adjustment range of the heater power.

[0107] For example, methods for single crystal growth that primarily adjust the heating power and secondarily adjust the liquid outlet distance include: calculating the corresponding time F′ h (t), in |F′ h (t) / (A h *ω h When F′|≥ set value x: ① If F′ h (t)>0, meaning F′ h (t) / (A h *ω h If F' ≥ the set value x, then increase the actual liquid outlet distance while increasing the heating power; ② If F′ h (t) < 0, meaning F′ h (t) / (A h *ω h )<0 and |F′ h (t) / (A h *ω h If |≥ set value x, then the actual liquid outlet distance is reduced while reducing the heating power; ③ If F′ h (t) < 0, meaning F′ h (t) / (A h *ω h )<0 and |F′ h (t) / (A h *ω h If |≥ set value x, then the actual liquid outlet distance will be reduced while reducing the heating power; ④ If F′ h (t)>0, meaning F′ h (t) / (A h *ω h If the value is greater than or equal to the set value x, then the actual liquid outlet distance will be increased while increasing the heating power.

[0108] Furthermore, point ① above also includes the premise F. h If (t) > d, then increase the heating power and increase the actual liquid outlet distance; point ② above also includes the premise F h If (t) > d, then reduce the heating power and the actual liquid outlet distance; point ③ above also includes the premise F h If (t) < d, then reduce the heating power and the actual liquid outlet distance; point ④ above also includes the premise F hIf (t) < d, then increase the heating power and increase the actual liquid outlet distance.

[0109] For example, methods for single crystal growth that primarily adjust the liquid outlet distance and secondarily adjust the heating power include: calculating the corresponding time F′ h (t), in |F′ h (t) / (A h *ω h When F < set value x: ① If F h (t)>d, and F′ h (t)>0, then increase the actual liquid outlet distance while increasing the heating power; ②If F h (t)>d, and F′ h (t) < 0, then reduce the actual liquid outlet distance while reducing the heating power; ③ If F h (t) < d, and F′ h (t) < 0, then reduce the actual liquid outlet distance while reducing the heating power; ④ If F h (t) < d, and F′ h If (t) > 0, then the actual liquid outlet distance should be increased while increasing the heating power.

[0110] In some embodiments, the change in the crucible's rising rate is Δv, where Δv = |k*ΔD / Δt|, ΔD is the difference in the diameter of the crystal rod between two adjacent measurement moments, and Δt is the time interval between two adjacent measurement moments, for example, ΔD / Δt = (D z+1 -D z ) / (t z+1 -t z ), where 0.1≤k≤2. Therefore, the rate of change of the crucible rising rate is positively correlated with |ΔD / Δt|, where |ΔD / Δt| is the absolute value of the ratio of the difference between the actual diameters of two adjacent measurements of the crystal rod to the time interval between the corresponding two adjacent measurement moments. That is, the greater the difference between the actual diameters of two adjacent measurements of the crystal rod, the greater the change in the crucible rising rate, facilitating timely matching and control measures for the crystal rod diameter.

[0111] For example, k can be 0.1, 0.5, 0.8, 1, 1.3, 1.6, 1.8, or 2, etc.

[0112] In some embodiments, the actual gate distance fluctuates within the range of -5mm to 5mm, with the target gate distance as zero. Setting the target gate distance is a process requirement in the crystal rod preparation process, ensuring the relative stability of factors such as temperature gradient and melt during crystal growth. The target gate distance needs to be maintained within a relatively stable range. When controlling the gate distance by changing the crucible rising speed, it is necessary to ensure that the actual gate distance fluctuates within the target gate distance range, without excessively affecting the V / G value, thereby achieving the stability of the crystal rod growth process and improving production efficiency. In other words, the process of reducing the actual gate distance in the above scheme will not reduce the actual gate distance to a value less than the minimum actual gate distance; similarly, the process of increasing the actual gate distance will not increase the actual gate distance to a value exceeding the maximum actual gate distance.

[0113] It is understandable that, based on the fluctuation of the actual diameter of the crystal rod, the adjustment of the actual liquid nozzle distance will also fluctuate to a certain extent. Therefore, in the two adjacent processes of increasing and decreasing the actual liquid nozzle distance, regardless of the order of the two processes, the actual liquid nozzle distance will tend to be adjusted to the target liquid nozzle distance when the two processes are completed.

[0114] The control device for a single crystal growth apparatus according to an embodiment of this application includes: a measuring mechanism, a data processing mechanism, a judgment mechanism, and a power adjustment mechanism. The measuring mechanism is used to measure the actual diameter D of the crystal rod, and can measure the actual diameter D of the crystal rod in real time and accurately, providing basic data for subsequent data processing and judgment.

[0115] The data processing mechanism communicates with the measurement mechanism, and the data processing mechanism is used to construct an initial function D = F0(t) between the actual diameter of the crystal rod and time t, where F0(t) = A0*sin(ω0*t+Φ)+d, or F0(t) = A0*cos(ω0*t+Φ)+d, d is the target diameter of the crystal rod, A0 is the initial amplitude, ω0 is the initial angular frequency, and Φ is the initial phase, which can intuitively reflect the changing trend of the crystal rod diameter with time.

[0116] The judgment mechanism determines whether F0(t) meets the preset conditions. When F0(t) meets the preset conditions, the crystal pulling rate is locked. The data processing mechanism sequentially uses the measurement results of the actual diameter of the crystal rod at n different measurement times to perform piecewise correction on the initial function, so as to obtain the piecewise function D = F0(t) between the actual diameter of the crystal rod and time t. h (t), F h (t)=A h *sin(ω h *t+Φ)+d, or, F h (t)=A h *cos(ω h *t+Φ)+d, where A h The corrected amplitude, ωh The corrected angular frequency is D = F. h (t) includes several sub-functions: the function between the actual diameter of the crystal rod at time t0 and time t is the initial function D = F0(t); the calibrated function D = F1(t) at time t1; and t z The time after calibration is D=F z (t), t z The time is the z-th measurement time after the preset conditions are met, where z is a positive integer and 2≤z≤n; the data processing mechanism is used to ensure that, in two adjacent measurements, one of the parameters corresponding to the amplitude and angular frequency of the latter measurement is equal to the corresponding parameter of the former measurement, and the other is calculated to achieve a correction.

[0117] The power adjustment mechanism is used to adjust the heating power, and the judgment mechanism is also used to judge F′. h (t) / (A h *ω h The relationship between F′ and the set value x, if |F′ h (t) / (A h *ω h If F′ ≥ the set value x, the power adjustment mechanism increases the heating power. h (t) / (A h *ω h )<0 and |F′ h (t) / (A h *ω h If |≥ set value x, then the power adjustment mechanism reduces the heating power, by judging F′ h (t) / (A h *ω h By precisely adjusting the heating power according to the relationship between the actual diameter of the crystal rod and the set value x, the heating power can be precisely controlled, which helps to make the actual diameter of the crystal rod fluctuate towards the target diameter. This makes the change in the actual diameter of the crystal rod more stable and the fluctuation period shorter during the constant diameter stage in the actual production process, which helps to improve the accuracy of controlling the actual diameter of the crystal rod.

[0118] In some embodiments, the data processing mechanism sequentially uses the measurement results of the actual diameter of the crystal rod at n different measurement times to perform piecewise correction on the initial function. In this case, the data processing mechanism is configured such that: when F0(t) meets the preset condition, the first measurement of the actual diameter D1 of the crystal rod at time t1 corresponds to an amplitude A1 equal to the initial amplitude A0, and the angular frequency ω1 is calculated to achieve a first correction; or, the judgment mechanism is configured such that: when F0(t) meets the preset condition, the first measurement of the actual diameter D1 of the crystal rod at time t1 corresponds to an angular frequency ω1 equal to the initial angular frequency ω0, and the amplitude A1 is calculated to achieve a first correction.

[0119] In some embodiments, the data processing mechanism is configured such that: in p consecutive corrections, at least one correction calculates the parameter amplitude, and at least one correction calculates the parameter angular frequency, where p is a positive integer and p≥2.

[0120] In some embodiments, the judgment mechanism is used to judge whether F0(t) meets a preset condition. When F0(t) = A0*sin(ω0*t + Φ) + d, the preset condition is F0(t) = d and F′0(t) > 0; when F0(t) = A0*cos(ω0*t + Φ) + d, the preset condition is F0(t) = d + A0. In some embodiments, after the judgment mechanism judges that F0(t) meets the preset condition, the measurement mechanism performs a measurement every preset time, and the preset time is less than 1 min.

[0121] In some embodiments, the data processing mechanism is configured such that: in any two adjacent corrections, one correction calculates the parameter amplitude, and the other correction calculates the parameter angular frequency.

[0122] In some embodiments, if the judgment mechanism judges that F′ h (t) / (A h *ω h )≥ the set value x, the power adjustment mechanism increases the heating power by an increment of ΔP1. If the judgment mechanism judges that F′ h (t) / (A h *ω h )<0 and |F′ h (t) / (A h *ω h )|≥ the set value x, the power adjustment mechanism decreases the heating power by a decrement of ΔP2; the power adjustment mechanism is configured such that: ΔP1 = m1*|F′ h (t) / (A h *ω h )|, ΔP2 = m2*|F′ h (t) / (A h *ω h )|, where m1 and m2 are both temperature control coefficients, 0 < m1 ≤ 15, 0 < m2 ≤ 15. For example, both m1 and m2 are positively correlated with ΔD, where ΔD is the difference in the diameter of the crystal bar at two adjacent measurement times.

[0123] In some embodiments, the power adjustment mechanism is configured such that: when the heater power (i.e., the heating power) needs to be adjusted, the basic heating power of the heater for each adjustment is P0; the control device is configured such that: when the power adjustment structure adjusts the heater power, the data processing mechanism calculates the period T h of D = F h (t) as T h = 2π / ωh If the time is ≥40 min, then the basic heating power P0 of the calibrated heater is made such that the corrected P0 equals t. h The average heating power over a preset time period prior to the specified time, where the preset time ranges from T. h / 8~T h / 2; if T h If the time is less than 40 minutes, the P0 value remains unchanged from the previous calibration.

[0124] In some embodiments, the control device further includes a drive mechanism for driving the crucible to rise and fall. The drive mechanism communicates with a judgment mechanism to control the rise and fall of the crucible according to the judgment mechanism. At this time, the judgment mechanism determines F′. h (t) / (A h *ω h When ) < set value x: If the judgment mechanism further determines F h (t)>d, and F′ h If (t) > 0, increase the actual liquid outlet distance; if F h (t)>d, and F′ h If (t) < 0, decrease the actual liquid outlet distance; if F h (t) < d, and F′ h If (t) < 0, decrease the actual liquid outlet distance; if F h (t) < d, and F′ h (t)>0, increase the actual liquid outlet distance.

[0125] In some embodiments, the driving mechanism is configured such that the change in the crucible's rising rate is Δv, where Δv = |k*ΔD / Δt|, ΔD is the difference in the diameter of the crystal rod between two adjacent measurement moments, and Δt is the time interval between two adjacent measurement moments, where 0.1 ≤ k ≤ 2. In some embodiments, the driving mechanism is configured such that it drives the crucible to rise and fall to change the actual liquid nozzle distance, and the actual liquid nozzle distance fluctuates within the range of -5 mm to 5 mm, with the target liquid nozzle distance as the zero point.

[0126] The single crystal growth apparatus according to an embodiment of this application includes: a furnace body, a crucible, a heater, and a control mechanism. The furnace body provides a stable and enclosed environment for crystal rod growth, which helps to form a stable thermal field and reduces the impact of temperature fluctuations, impurity contamination, etc., on the crystal rod growth process. The crucible is located inside the furnace body and defines a holding space. The crucible is used to hold the raw materials for crystal rod growth. By adjusting parameters such as the position, rotation speed, and lifting speed of the crucible, the temperature gradient, melt conditions, etc., during the crystal rod growth process can be controlled, thereby affecting the quality of the crystal rod. The heater is located inside the furnace body and is used to heat the crucible. It is the main heat source in the crystal rod growth process. By heating the raw materials in the crucible to a molten state, it provides the necessary conditions for crystal rod growth. The control device is the control device of the crystal rod growth apparatus according to the above embodiment of this application. The power adjustment mechanism is used to adjust the power of the heater. The measuring mechanism monitors parameters such as the actual diameter of the crystal rod in real time and transmits the data to the data processing mechanism for analysis and processing, which helps to achieve precise control of the crystal rod growth process. These mechanisms are set in the crystal rod growth apparatus and each plays an important role, jointly realizing the precise control and optimization of the crystal rod growth process.

[0127] In the description of this specification, references to terms such as "one embodiment," "another embodiment," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment is included in at least one embodiment of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples, without contradiction. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

Claims

1. A method for single crystal growth, wherein, include: During the constant diameter growth stage, the initial function between the actual diameter of the crystal rod and time t is denoted as D=F0(t), F0(t)=A0*sin(ω0*t+Φ)+d, or F0(t)=A0*cos(ω0*t+Φ)+d, where d is the target diameter of the crystal rod, A0 is the initial amplitude, ω0 is the initial angular frequency, and Φ is the initial phase; When F0(t) meets the preset condition, the crystal pulling rate is locked, and the initial function is piecewise corrected using the measurement results of the actual diameter of the crystal rod at n different measurement times, so as to obtain the piecewise function D = F0(t) between the actual diameter of the crystal rod and time t. h (t), F h (t)=A h *sin(ω h *t+Φ)+d, or, F h (t)=A h *cos(ω h *t+Φ)+d, where A h The corrected amplitude, ω h The corrected angular frequency is D = F. h (t) includes several sub-functions: the function between the actual diameter of the crystal rod at time t0 and time t is the initial function D = F0(t); the calibrated function D = F1(t) at time t1; and t z The function D = F after time calibration z (t), t z The time is the z-th measurement time after the preset conditions are met, where z is a positive integer and 2≤z≤n; Calculate F at the corresponding time. h '(t), if F h '(t) / (A h *ω h If F ≥ the set value x, then increase the heating power; if F h '(t) / (A h *ω h ) < 0 and |F h '(t) / (A h *ω h If |≥ set value x, then reduce the heating power, where A h ω h t h The amplitude and angular frequency at time x > 0 In two consecutive measurements, one of the parameters corresponding to the amplitude and angular frequency of the latter measurement is equal to the corresponding parameter of the former measurement, while the other is calculated to achieve a correction.

2. The method for single crystal growth according to claim 1, wherein, The step of sequentially using the measurement results of the actual diameter of the crystal rod at n different measurement times to perform piecewise correction on the initial function includes: When F0(t) meets the preset conditions, the actual diameter D1 of the crystal rod is measured for the first time at time t1. The corresponding parameter amplitude A1 is equal to the initial amplitude A0, and ω1 is calculated to achieve a first correction; or, The corresponding parameter amplitude ω1 is equal to the initial angular frequency ω0, and A1 is calculated to achieve a first correction.

3. The method for single crystal growth according to claim 1 or 2, wherein, In the p consecutive corrections, at least one correction calculates the parameter amplitude, and at least one correction calculates the parameter angular frequency, where p is a positive integer and p≥2.

4. The method for single crystal growth according to claim 3, wherein, In any two consecutive corrections, one correction calculates the amplitude of the parameter, and the other correction calculates the angular frequency of the parameter.

5. The method for single crystal growth according to any one of claims 1-4, wherein, The constant diameter growth stage includes multiple sub-stages that proceed sequentially over time. The set value x corresponding to the multiple sub-stages is not equal, and the set value x of the multiple sub-stages decreases in chronological order.

6. The method for single crystal growth according to claim 5, wherein, The current constant diameter length of the crystal rod is L, the target constant diameter length of the crystal rod is L', and the following conditions are met: When 100mm≤L<300mm, x=x1, When 300mm≤L<L'-500mm, x=x2. When L'-500mm≤L≤L', x=x3, 0.01≤x3≤1 / 2, where L'≥1300mm.

7. The method for single crystal growth according to claim 6, wherein, When 300mm≤L<L' / 2 When L' / 2 ≤ L < L' - 500 mm 8. The method for single crystal growth according to any one of claims 1-7, wherein, When F0(t) = A0*sin(ω0*t+Φ)+d, the preset condition is that F0(t) = d and F0'(t) > 0; When F0(t) = A0*cos(ω0*t+Φ)+d, the preset condition is F0(t) = d+A0.

9. The method for single crystal growth according to any one of claims 1-8, wherein, Once F0(t) meets the preset conditions, a measurement is performed every preset time interval, which is less than 1 minute.

10. The method for single crystal growth according to any one of claims 1-9, wherein, If F h '(t) / (A h *ω h If the value is greater than or equal to the set value x, then the heating power is increased by ΔP1. If F h '(t) / (A h *ω h ) < 0 and |F h '(t) / (A h *ω h If |≥ set value x, then the heating power is reduced by ΔP2, where, ΔP1 = m1 * |F h '(t) / (A h * ω h )|, ΔP2 = m2 * |F h '(t) / (A h * ω h )|, where m1 and m2 are both temperature control coefficients, 0 < m1 ≤ 15, 0 < m2 ≤ 15, and m1 and m2 are both positively correlated with ΔD, where ΔD is the difference in the diameter of the crystal bar at two adjacent measurement times.

11. The method for single crystal growth according to claim 10, wherein, When the heater power needs to be adjusted, the base heating power of the heater for each adjustment is P0, and D = F is calculated. h The period T of (t) h =2π / ω h , If T h If the time is ≥40 min, then the basic heating power P0 of the calibrated heater is made such that the corrected P0 equals t. h The average heating power over a preset time period prior to the specified time, where the preset time ranges from T. h / 8~T h / 2; If T h If the time is less than 40 minutes, the P0 value remains unchanged from the previous calibration.

12. The method for single crystal growth according to claim 10 or 11, wherein, The heating power is changed by the top heater on the upper side of the crucible, where m1≤1 and m2≤1; or, The heating power is changed by the side heater on the outer periphery of the crucible, where 1≤m1≤5 and 1≤m2≤5; or, The heating power is changed by the bottom heater on the lower side of the crucible, 5≤m1≤15, 5≤m2≤15.

13. The method for single crystal growth according to any one of claims 1-12, wherein, |F h '(t) / (A h *ω h When | < set value x, If F h (t)>d, and F h If '(t)>0, increase the actual liquid outlet distance; if F h (t)>d, and F h If '(t) < 0, decrease the actual nozzle distance; if F h (t) < d, and F h If '(t) < 0, decrease the actual nozzle distance; if F h (t) < d, and F h If '(t)>0, increase the actual liquid outlet distance.

14. The method for single crystal growth according to claim 13, wherein, The change in the crucible's rising rate is Δv, where Δv = |k*ΔD / Δt|, ΔD is the difference in the diameter of the crystal rod between two adjacent measurement moments, and Δt is the time interval between two adjacent measurement moments, where 0.1 ≤ k ≤ 2.

15. The method for single crystal growth according to claim 13 or 14, wherein, The actual liquid outlet distance is taken as the target liquid outlet distance and fluctuates within the range of -5mm to 5mm.

16. A control device for a single crystal growth apparatus, wherein, include: The measuring mechanism is used to measure the actual diameter D of the crystal rod; The data processing mechanism communicates with the measuring mechanism and is used to simulate the initial function D = F0(t) between the actual diameter of the crystal rod and time t, where F0(t) = A0*sin(ω0*t+Φ)+d, or F0(t) = A0*cos(ω0*t+Φ)+d, d is the target diameter of the crystal rod, A0 is the initial amplitude, ω0 is the initial angular frequency, and Φ is the initial phase; The judging mechanism is used to determine whether F0(t) meets a preset condition. When F0(t) meets the preset condition, the crystal pulling rate is locked, and the initial function is piecewise corrected using the measurement results of the actual diameter of the crystal rod at n different measurement times, so as to obtain the piecewise function D = F0(t) between the corrected actual diameter of the crystal rod and time t. h (t), F h (t)=A h *sin(ω h *t+Φ)+d, or, F h (t)=A h *cos(ω h *t+Φ)+d, where A h The corrected amplitude, ω h The corrected angular frequency is D = F. h (t) includes several sub-functions: the function between the actual diameter of the crystal rod at time t0 and time t is the initial function D = F0(t); the calibrated function D = F1(t) at time t1; and t z The function D = F after time calibration z (t), t z The time is the z-th measurement time after the preset conditions are met, where z is a positive integer and 2≤z≤n. The data processing mechanism is used to ensure that, in two adjacent measurements, one of the parameters corresponding to the amplitude and angular frequency of the latter measurement is equal to the corresponding parameter of the former measurement, and the other is calculated to achieve a correction. The power adjustment mechanism is used to adjust the heating power, and the judgment mechanism is also used to judge F. h '(t) / (A h *ω h The relationship between F and the set value x, if F h '(t) / (A h *ω h If F ≥ the set value x, then the power adjustment mechanism increases the heating power. h '(t) / (A h *ω h ) < 0 and |F h '(t) / (A h *ω h If |≥ set value x, then the power adjustment mechanism reduces the heating power.

17. A single crystal growth apparatus, wherein, include: A furnace body and a crucible, wherein the crucible is disposed within the furnace body and defines a holding space; A heater, which is disposed inside the furnace and is used to heat the crucible; The control device is the control device for the single crystal growth equipment according to claim 16, and the power adjustment mechanism is used to adjust the power of the heater.

Images